Incorporating Human and Natural Adaptations in Assessing Climate Change Impacts on Wildfire Occurrence pp. 61-73
Authors: Jianbang Gan
Abstract: Human and natural adaptations to climate change are evident, yet incorporating these adaptations into climate change impact assessments has been a challenging task. This study applies the panel data modeling approach to examining the impact of global climate change on the occurrence of human- and nature-caused wildfires in the US. Estimated using actually observed wildfire data in the 48 continental states of the US from 1991-1997, the panel data models are able to incorporate human and natural adaptations into the impact assessment as the data reflect changes in the values of relevant variables along the temperature and precipitation gradients. In addition, the panel data models are better able to control for the effect of missing or unobserved variables and to ease the possible multicollinearity problem associated with highly correlated climate variables. The results indicate that the number of human-caused wildfires would increase with increases in spring temperature and decreases in precipitation in all seasons, and that nature-caused wildfire incidents would rise with increases in winter and summer temperature and decreases in summer precipitation. Based on the temperature and precipitation changes predicted by the Hadley Centre and Canadian General Circulation Models under a 2◊CO2 climate, the number of nature-caused wildfires would increase by about 2.5 times, and that of human-caused wildfires would change from -7% to +37%. Thus, if the predicted climate change occurs, it would tremendously intensify the occurrence of nature-caused wildfires, creating challenges for wildfire protection and mitigation.